"Exploring History in the Wargame" Topic

Phil:
That is great to hear. Do you feel I understand what you were saying? I don't think we were taking uncompatible positions.

Timbo:
Your welcome. You're right, it would take some effort, but probably no more than designers put into 'playtesting' the game aspects at the moment.

Let me give a more specific example of those tests, one that demonstrates how a wargame can provide the historical recognition visa vie a game noted and valued at the beginning of the Thread.

One Turing Test that I like would be to take up to 20 or 30 players experienced playing the rules being validated. Then the designer targets particular mechanics see whether the players can connect the game play to actual historical dynamics.

For instance, Le Sacre Feu has brigades move using regulating battalions. Experienced players with the system are given several historical incidents relating the the dynamics of moving brigades, then not being familiar the actual historical events, they are asked to infer conditions or predict what will happen.

One example could be Leith's division advancing at Salamanca. In the advance, Leith sent majors to each end of the first brigade to keep the line straight. Knowing that, could the players guess where Leith stationed himself and where the regulating battalion was in the line?

Or Benning's brigade on the 2nd day of Gettysburg. He had to go to the left of the line to get them realigned with the rest of the brigade. Where was the regulating battalion for his brigade? And knowing where the regulating battalions were placed, what direction was Leith's and Benning's advance in comparision to the main battleline?

If LSF accurately portrays the dynamics of moving brigades with regulating battalions then the majority of the players will be able to answer the questions correctly, without knowing the actual history.

IF that is the case, then the designer can say with a great deal of confidence that the mechanics for simulating regulating battalions do mimic history AND that future players can and probably will experience that recognition between the game system and historical narratives so valued by wargamers.

No fuzziness, no guessing. This is one of my favorite tests because of the 'Ah has' players and I had with the tests and debriefings. That type of testing can be done for separate subsystems of the game design and for the system as a whole.

With four different kinds of tests validating the design in such a fashion, the designer can say with confidence--because it's been proven:

1. Yes, the game system does simulate the history selected by the designer.

2. Players can read history and see the connections between it and playing the wargame.

3. The game system will simulate those selected parts of history faithfully regardless of the scenario or odd play.

The designer gets to chose what history he wants to simulate, whatever he wants.

He choses what game mechanics to include and how they work to portray history, using all the art, science and subtlety at his disposal.

And then he tests to see if his creation actually does simulate those elements of history he decided to portray.

Bill H.

My dyslexia has struck again. That should be

Le Feu Sacré and LFS

*blush*

One Turing Test that I like would be to take up to 20 or 30 players experienced playing the rules being validated. Then the designer targets particular mechanics see whether the players can connect the game play to actual historical dynamics.

Um mm……..Turing Test? As in Alan Turing? Are you using this test in the correct context?

Bowman:

Yes and no. The original "Turing Test" dealt with computer systems and artificial intelligence, thought up by Alan Turing, as you note. The idea was that a computer could be said to "think" if a human interrogator could not tell it apart, through conversation, from a human being. The same comparison test has been used in systems testing in a number of ways since then, as well as simulations, but it is still called a 'Turing Test'.

In a dynamic, participatory simulation parlance, the 'Turing Test' says if a person can't tell the difference between the simulation and the real thing [in some specfic respect] the design does simulate at that point. This can be done in reverse, where the person, having only experience with the simulation, can then respond correctly to similar real world activities or challenges.

Osman Balci, who contributed the chapter on Verification, Validation, and Testing in Handbook of Simulation describes the Turing Test for Simulations this way:

Turing Test: The Turing test is based on the expert knowledge of people about the system under stud. The experts are presented with two sets of output data obtained, one from the model and one from the [real] system, under the same input conditions. Without identifying which one is which, the experts are asked to differentiate between the two. If they succeed, they are asked how they were able to do it. Their response provides valuable feedback fo rcorrecting the model representation. If they cannot differentiate, our confidence in model validity is increased.

Osman is referring to computer system outputs, but the Turing test can be used with any kind of simulation, either with subsystems or the entire design, processes or outputs.

Is that clear?

Bill H.

Thanks Bill,

Kind of. It seems that the "Simulation" community has appropriated the term and uses it differently than originally warranted. That's why I asked.

So looking at your statement:

In a dynamic, participatory simulation parlance, the 'Turing Test' says if a person can't tell the difference between the simulation and the real thing [in some specfic respect] the design does simulate at that point. This can be done in reverse, where the person, having only experience with the simulation, can then respond correctly to similar real world activities or challenges.

Let's deal with the second sentence first. Flight simulators would fall into that category as any one could see.

The first sentence and Balci's quote intrigue me. Let me see if I can apply this to my wargaming. Let's use WW2 as I have been using that as my examples. My friends are over and we will "play" a Saipan scenario using Disposable Heroes (my current game of choice).

So we load up our scenario with the requisite data. We know that US Marines and the IJA were the belligerents, the size and nature of their infantry squads, the type of support each side had, and the type of armor they would use. Already the "exploring history in the Wargame" concept is moot as I had to have sufficient reading ahead of time to get my scenario right.

I will take the Japanese side. OK, first big fail of a Turing Test. There are no 600 foot godlike General who can command one side and see the entire battlefield, and knows exactly what every unit is doing at any given time. My opponents sees that I have brought out my Japanese Tank collection from my cabinet, and is already thinking that he should have a bazooka team at any position that provides a clean line of sight. A big disconnect from reality.

Ok, the ranges involved with movement and shooting range have no basis in reality. In every game I play these are tweaked so that we get all our pieces on a reasonably sized table within arms reach of the participants. It is a suspension of disbelief that we all have to deal with. Again a failed Turing Test.

My Japanese squads maneuvre into position, by suppression by the LMG squad fire or by small Type 89 Mortar fire. This way the rifle squad moves up into position, to then offer suppressing fire to bring up the other squads. This works well when you have sufficient coverage, but it still requires some wargaming skill. When I get slaughtered in this game it is usually against someone who knows how to move his troops efficiently, with a minimum of casualties and then be able to bring a concentrated fire upon my troops. I like this system and "think" it mimics movement quite well. A Turing Test pass.

Does this simulate what the soldiers are feeling, being ordered to advance against a hail of counter-fire and artillery fire? Not in the least. Turing Test fail.

I maneuvre my Type 92 HMG into the lead squad of Marines, who are within Banzai charge range. My Type 92 fires at a greater rate than normal gun fire, and being hit by it causes a greater Morale check than regular fire. This seems to mimic the power of a heavy machine gun on the battlefield. Therefore Turing Test pass.

However, the Type 92 tended to jam a lot. No rules for that. Failed Turing Test

The Type 92 didn't have the same rate of fire as the US 30 calibre M1917. Therefore, the Japanese player rolls 4 die for his shooting as opposed to 5 die for the American gun. Ok. I can buy that. Passed Turing Test.

The fact that we roll die, big time Turing Test fail. We randomize outcomes according to a % valuation. This is a necessary compromise, but a total disconnect from reality. The "fact" that the Japanese HMG fires about 10% less effectively than a Browning 30 cal is because that represents one PIP on a D10 die.

Then time for my Tank charge. The Marines respond only once the tanks are in sight. Why? Becuase, real soldiers would have heard them and felt the ground rumble well ahead of seeing them. Turing Test fail.

My Tanks die in droves! Turing Test pass!

It seems that for every "connect" our game has with reality, it is just as easy to point out a "disconnect". Based on the simple Turing Test, this would produce a low level of simulation. However, to be fair, each "connect" and "disconnect" are not weighted evenly. I'll accept that without the "600 foot general" throwing dice, I wouldn"t be there having fun. I'll overlook the gaminess of having Die PIPS determine the granularity of outcomes, if other more important mechanisms "seem" real.

I'm truly sorry, but I am still not convinced that these wargames are such accurate simulations that we can glean historical insights from them. For me, that was done in the reading before the wargame, not during the execution of the wargame.

If I have this totally wrong, I apologize in advance and look forward to hearing from you.

Your mileage may undoubtably vary.

Bowman wrote:

Kind of. It seems that the "Simulation" community has appropriated the term and uses it differently than originally warranted. That's why I asked.

Bowman:
Actually, the simulation community is only one of several system analysis 'communities' that have appropriated the term, but yes they have taken the original IA application and used the comparison technique elsewhere.

I appreciate the application example you provided. It is a good one. However, in your application of a 'Turing Test', you have tried to test too much with too loose an approach without a 3rd party. And unfortunately, like most all wargamers, without any concrete information about what the wargame was designed to simulate. And I apologize for the length of this, but I wanted to address each of your comments.

The Turing Test explanation: In a dynamic, participatory simulation parlance, the 'Turing Test' says if a person can't tell the difference between the simulation and the real thing [in some specfic respect] the design does simulate at that point. This can be done in reverse, where the person, having only experience with the simulation, can then respond correctly to similar real world activities or challenges.

Let's deal with the second sentence first. Flight simulators would fall into that category as any one could see.

The first sentence and Balci's quote intrigue me. Let me see if I can apply this to my wargaming. Let's use WW2 as I have been using that as my examples. My friends are over and we will "play" a Saipan scenario using Disposable Heroes (my current game of choice).

Awful title. Unfortunately, Balci was describing an entire process, and you only acted on the first sentence…

So we load up our scenario with the requisite data. We know that US Marines and the IJA were the belligerents, the size and nature of their infantry squads, the type of support each side had, and the type of armor they would use. Already the "exploring history in the Wargame" concept is moot as I had to have sufficient reading ahead of time to get my scenario right.

In a turing test, you aren't 'exploring history in a wargame'--that is another issue. In a turing test, the participants aren't the ones doing the test, they are generating the results to be tested. Big difference. In a reverse test, after the participants play the game, they are exposed to history they don't know. Either way, your approach is internal and experiential, not a turing test.

I will take the Japanese side. OK, first big fail of a Turing Test. There are no 600 foot godlike General who can command one side and see the entire battlefield, and knows exactly what every unit is doing at any given time. My opponents sees that I have brought out my Japanese Tank collection from my cabinet, and is already thinking that he should have a bazooka team at any position that provides a clean line of sight. A big disconnect from reality.

Again, is that what the Turing test is targeting, the 600 foot general syndrome? And yes, YOU obviously have a great deal more information than the original combatants. It is a big disconnect from reality…at that point. There is nothing that can be done about the knowledge you possess unless you change the scenario and troop counts etc. Which could be done. But again, what exactly are you targeting with the Turing Test?

Ok, the ranges involved with movement and shooting range have no basis in reality.

None? By design? If so, there's no point in testing them.

In every game I play these are tweaked so that we get all our pieces on a reasonably sized table within arms reach of the participants. It is a suspension of disbelief that we all have to deal with. Again a failed Turing Test.

You haven't said what you are testing, and again, the Turing Test requires review from folks OUTSIDE the simulation experience, or without knowledge of the history being simulated. Just an aside here though: All simulations, regardless of purpose or design, require a suspension of belief from the participants or user to function, to be of any value.

My Japanese squads maneuvre into position, by suppression by the LMG squad fire or by small Type 89 Mortar fire. This way the rifle squad moves up into position, to then offer suppressing fire to bring up the other squads. This works well when you have sufficient coverage, but it still requires some wargaming skill. When I get slaughtered in this game it is usually against someone who knows how to move his troops efficiently, with a minimum of casualties and then be able to bring a concentrated fire upon my troops. I like this system and "think" it mimics movement quite well. A Turing Test pass.

That isn't a Turing test because you, the participant, thinks it mimics movement. If you documented the movement and/or results of the game for that particular action, and produced a like-documented historical movement and results, then gave it to a 3rd party as knowledgeable as you, could they, would they be able to pick the historical movement from the game movement? That is a turing test.

Does this simulate what the soldiers are feeling, being ordered to advance against a hail of counter-fire and artillery fire? Not in the least. Turing Test fail.

Was the simulation designed make the players feel they are being ordered to advance against a hail of counter-fire and artillery? No? Then, of course, it can't be tested for that. ALL simulations have a limit to how much they can simulate. Designers have to select what will and won't be simulated. A turing test is only valuable in testing what the simulation is designed to represent. Any test of the vast portion of reality not included in the design is a foregone conclusion--a failure.

And of coures, if you don't know what the designer selected to simulate, no test of it is possible, other than a shot-gun guessing game kind of affair by feel--which is exactly what you are doing. And you can't know whether the successes or failures of your 'test' have anything to do with what the wargame was designed to simulate.

I maneuvre my Type 92 HMG into the lead squad of Marines, who are within Banzai charge range. My Type 92 fires at a greater rate than normal gun fire, and being hit by it causes a greater Morale check than regular fire. This seems to mimic the power of a heavy machine gun on the battlefield. Therefore Turing Test pass.

Again, that isn't a turing test. You are playing the game and where the game meets your criteria for a pass, it does. You don't know if that is what other knowledgeable folks would conclude or even if the designer had that result as a design goal…unless he told you, which I surmise he hasn't.

However, the Type 92 tended to jam a lot. No rules for that. Failed Turing Test

There is no point in testing it unless the simulation was designed to cover that little detail. Maybe the designer didn't think that was important as you do or had information different from you, or found through some research that such jams didn't change combat outcomes, or maybe he was simply unaware of that altogether.

Do you know the answers to those questions? Probably not, so how can you test for it? Any 'test' is simply whether the game fits your view or not. A Bowman test, which is not anything like the eight I listed. There is nothing particularly objective or 3rd party about your test.

There's nothing wrong with that, but it won't establish objectively whether the simulation does what it was designed to do.

The Type 92 didn't have the same rate of fire as the US 30 calibre M1917. Therefore, the Japanese player rolls 4 die for his shooting as opposed to 5 die for the American gun. Ok. I can buy that. Passed Turing Test.

Really? So you know that four dice compared to 5 dice produces results that compare with historical results? Have you given that assumption a turing test with 3rd party experts?

The fact that we roll die, big time Turing Test fail. We randomize outcomes according to a % valuation. This is a necessary compromise, but a total disconnect from reality. The "fact" that the Japanese HMG fires about 10% less effectively than a Browning 30 cal is because that represents one PIP on a D10 die.

That isn't a disconnect from reality. The designer didn't design the game with the idea that die rolling was a part of WWII combat. It is a mechanic which supposedly produces an outcome that mimics the effectiveness of the HMG fire. The question to be tested is "Does it?" That is a testable question.

An astrophysicist uses a computer to simulate the behavior of a galaxy. Because there are silicon chips involved, is that a "disconnect from reality?" No, because the chips were only supporting the the simulation system on the screen. They weren't what was being simulated. All simulations are artificial constructs…artificial: i.e. disconnected from reality.

It is what they produce that is the question. Does the simulation system simulate… not whether you are rolling a dice or not. The die roll 'combat results' are what is being simualted and that result can be tested against reality.

Then time for my Tank charge. The Marines respond only once the tanks are in sight. Why? Becuase, real soldiers would have heard them and felt the ground rumble well ahead of seeing them. Turing Test fail.

Again. Lots of things that a simulation can't simulate. The question is whether the aspects of reality that were selected to be simulated pass the test. Unless the rules specifically were designed to address that rumble… why test for it?

My Tanks die in droves! Turing Test pass!

Not a turing test. You would have to show your game results along with similar historical events to a 3rd party expert [and those more 'objective'. the more experts used, the more statistically objective the results.] and if they couldn't tell the difference between the results or how the action progressed, THEN it passes a turing test.

It seems that for every "connect" our game has with reality, it is just as easy to point out a "disconnect". Based on the simple Turing Test, this would produce a low level of simulation.

When a simulation can only simulate a small portion of reality effectively, it is very easy to idenitify the vast portions of reality not addressed, and thus any design fails to simulate.

The only points where you can fairly say a simulation suffers a 'disconnect' are those points the designer specifically tried to represent--and failed.

I don't know what constitutes a 'low level' of simulation for you, but I do know that you may well have been testing "Disposable Heros" for many things it was never designed to simulate in the first place. The problem is that the designer probably hasn't given you enough information to know the answers to that question.

However, to be fair, each "connect" and "disconnect" are not weighted evenly. I'll accept that without the "600 foot general" throwing dice, I wouldn"t be there having fun.

Terrific. So the many things that the desired '600 foot general" may keep the wargame from simulating aren't part of what can be or is simulated within that constraint.

I'll overlook the gaminess of having Die PIPS determine the granularity of outcomes, if other more important mechanisms "seem" real.

Your choice. Your choice as to what is 'more important' to you and your sense of what 'seems' real. The tests are to determine whether the mechanics the designer felt were important to include, do in fact, objectively, mimic real processes and environmental considerations.

I'm truly sorry, but I am still not convinced that these wargames are such accurate simulations that we can glean historical insights from them.

I never said that current simulations are accurate. I have said that I don't know if they are and nobody does because:

1. Designers haven't been upfront and clear about what they have SELECTED to simulate.

2. They haven't specified how the mechanics they have chosen are meant to simulate those selected aspects of reality, and

3. Because of that information vaccuum, there is no way to test to see if they are successful. Your 'seems' to be real is the 'feels right' subjective test of 'I like it' situation we have right now. Everyone knows what they know and what they like. No problem with that.

The problem is what designers know and are doing in their designs--what you were attempting to test for. I am simply pointing out that there are, currently, proven ways to establish in very concrete terms whether and where wargames do simulate.

For me, that was done in the reading before the wargame, not during the execution of the wargame. If I have this totally wrong, I apologize in advance and look forward to hearing from you.

No need to apologize. I have lived with this stuff for a long time, and know it works, so I often find it hard to disassemble, and explain to others who haven't dealt with it before.

You said, "but I am still not convinced that these wargames are such accurate simulations that we can glean historical insights from them."

I agree. At this point it is a guessing game as to what historical insights went into their design or how the design actually simulates. So it is not surprising that wargamers should be unsure about both historical accuracy, if any, or what historical insights could be gleaned from such mystery meat.

That lack of clarity is totally unnecessary and I believe detrimental to the wargame hobby.

Bill H.

So I guess in order to do the Turing test on Bowman's wargame, you could

- find an after-action report of the historical action,
- write up an AAR from the wargame in a similar style, but remaining faithful to the results of the wargame
- get an 'expert' (but not so expert as to recognise the specific 'real' AAR) to read the two and see if the 'fake' can be identified
- repeat to taste

Timbo:

Yes, but it tastes best with at least five of those experts reviewing the same data. And actually, AARs can be written in such a way as not to give away the story…or if the test is for a specific subsystem of the game, then just that information. There are forms for such things….

And remember that this is just one kind of test out of eight basic kinds and the rule of thumb is four different test 'passes' are necessary for reasonable confidence.

Bill

Bill let me take an example:

My Tanks die in droves! Turing Test pass!

Not a turing test. You would have to show your game results along with similar historical events to a 3rd party expert [and those more 'objective'. the more experts used, the more statistically objective the results.] and if they couldn't tell the difference between the results or how the action progressed, THEN it passes a turing test.

You don't need any expert. The Japanese tanks attacked and were wiped out, to a man. Same thing in Peleliu. The game ran exactly like historical results, so according to your comment above, it passes a Turing Test.

Now if my Marine playing friend rolled poorly and my tanks survived, then you don't need any experts again to tell that the games results are diametrically opposed to historical events. Is that a Turing Fail?

More importantly, how do these polarized results explore history (to naggingly get back to the actual topic).

Bowman:

You asked how a turing test works. If you decide you don't need an expert, fine, but that's not the point of a turing test. It asks if people well-versed in that history/combat tactics would come to the same conclusion you did, considering the number of such tank attacks in the Pacific. Those folks don't play the game to find out, they are outside of that experience and thus more objective about it than you. [That is the idea, which is why coaches are also judges at the Olympics, etc.] They may well come to the same conclusion you do, but you test it to make sure.

In the simulation, how, where, when and why those tanks were wiped out can and maybe be tested too. I can create a one D6 die-roll mechanic for Japanese tank attacks. 1-6 has the tanks die en mass. Is that a simulation, and if it proved to be one, would you care? The question is about how the simulation system works [a series of decisions and results]as well as any end results.

Now if my Marine playing friend rolled poorly and my tanks survived, then you don't need any experts again to tell that the games results are diametrically opposed to historical events. Is that a Turing Fail?

How can the test fail? You just decided it wasn't necessary. The questions needing to be asked in decigning the system would be such things as 'what historical events?.' Do all the recorded events have the same results? Why or why not? Those are questions supposedly experts would have the answers too, and considering the amount of history out there, maybe a different conclusion than yours.

More importantly, how do these polarized results explore history (to naggingly get back to the actual topic).

Do those polarized results have anything to do with actual history? A simulation is a dynamic system, that works by reproducing relationships between decisions and the historical situation. [Such as a Japanese player deciding to stage a mass tank attack. One historical option with historical consequences.]

A gamer can only explore the history selected by the designer to include in the game system. If it ain't there, you can't explore it. It is also hard to explore if you don't know what history was included, or what you are 'exploring.' That guessing game is evidence in several of the comments at the beginning of this thread.

One of the primary purposes of any simulation is to answer the 'what if' questions. [What if the Japanese dug the tanks in and made the Marines come them?]

Whenever I have seen anyone talk about 'exploring history' in a wargame, it has meant discovering the decision options available to the historical players and trying them out to see what happens. Gamers can explore the battlefield environment, discover why generals did what they did, often by trying something else and experiencing the dynamic results.

That can only happen within the parameters of the history selected by the designer. The tests establish that:

1. Yes, the mechanics do mimic the selected historical sources.
2. Yes, players do experience the same options, decisions and consequences of the historical generals in areas selected by the designer to simulate.
3. Yes, players can know in specific terms what history they can and can't explore with a particular design.[It has been proven by the tests] They won't bother to look for history that wasn't included [like jamming Japanese HMGs] and will know where the history is represented by what mechanics.
4. When they buy a wargame to 'explore history', they will hae some sense of what history is being provided for that exploration beyond knowing the scale and era represented.

No simulation can represent everything, nor will they ever be able to…which is a Good Thing. What gamers need to know is what history is being included, where it is represented, and whether the wargame successfully mimic that selected history.

It doesn't have to be guessing routine, "Gee, that seemed like the history I read, it 'seems' to work here, but not there, etc. etc. And why not? Who knows?

Bill H.

Bowman:

Just one more point. A simulation and/or wargame is a complex system, with a variety of subsystems that all have to work together to produce a play experience that mirrors history.

One of the things that has been found is that a designer can get all the parts right, like Japanese tank assaults being blown away, but the very mechanic that allows that, will screw up some other part of the system's ability to simulate. Tests were created in part because you can get all the data right and still have the system fail to do the job. I am sure you know of wargames where that apears happen.

Bill H.

I did it again:

[That is the idea, which is why coaches are also judges at the Olympics, etc.]

That should read "which is why coaches are NOT also judges at the Olympics.

A Turing test is done by outside observers for the same reason game designers blind-test their games… To get a more objective look at the game. A designer is too close to the design. A gamer playing the game, with their own views on what is right can be too.

Bill H.

I suppose the real test is simple. If I put one team of gamers vs a second team of gamers on a large patch of ground in Middle England each with …hmmm lets say 10,000 real troops to command armed with weapons from a historical period they regularly game and organised into formations of that period and using that periods communications and command systems, do they feel that as a result of playing wargames they will have a better chance of understanding what the hell is going on then an equivalent number of people plucked off the street (assuming the street isn't at Sandhurst, West Point etc). Lets also allow them the knowledge they have gleaned from doing research into tactics, weapons systems and such like as that is part of how we learn history from wargaming. what do you think?

Elenderil:

I think you've missed the point entirely. We are testing a simulation, an artificial system to see if it 'acts like' the data and sources of information the designer used to create his simulation.

All you'd be testing with your 'real test' is 10,000 real troops and how they feel, that is if you can find 'Middle Earth.'

I'm not avocating anything that isn't already done by thousands of simulation designers, using tests and methods that over several decades have been proven to work… against the 'real thing.' The whole point of a simulation is you don't have to go to the expense of arming 10,000 real troops to find out if a A particular simulation/wargame design does what the designer says it is designed to do: Simulate historical battles based on the information they chose to include.

And the nice thing about the 'experts' in our turing test is that they don't even have to know a thing about wargames. They are evaluating the history/combat dynamics generated by a particular set of rules against what they do know… the actual history, not other wargame rules.

Bill H.

No not missed the point entirely but rather as there appears to be an intractable point of disagreement proposing a "thought experiment".

If our random sample of gamers believe that they do have a better understanding then the next question in the experiment is why do you believe this to be the case. Has that better understanding arisen from background research or from the playing of wargames (or from having an ego the size of a small planet!).

I actually have a very clear understanding of the two viewpoints and was simply proposing a different approach to arrive at answer to the original question. My view remains that a gamer will acquire a better understanding of some aspects of history from a wargame. Other aspects will not be experienced from playing a wargame.

Wargames by their very nature cannot fully simulate all aspects of reality. A game designer takes a decision (consciously or not) as to where the game will focus. There is a clear tendancy too focus on aspects that can be clearly and relatively accurately modelled in the statistical modelling process that is a set of wargame rules. Equally well other aspects cannot be acurately modelled as their are simply too many variables for a set of pen and paper rule mechanisms to cover. \This is where the "fudge" factor has to be applied to best of the designers knowledge and belief, or what has been described above as a set of prejudices.

Bored Now.

My view remains that a gamer will acquire a better understanding of some aspects of history from a wargame. Other aspects will not be experienced from playing a wargame.

Elenderil:

I agree. My only point is the only 'better understanding of history' will be whatever the designer has successfully replicated with his wargame. If it ain't there, it can't be better understood. If the history is unsuccessfully portrayed in play, then obviously the gamer won't gain a better understanding.

Wargames by their very nature cannot fully simulate all aspects of reality. A game designer takes a decision (consciously or not) as to where the game will focus.

Absolutely, I agree. That is true of any and all simulations. So the question is what aspects of reality have been chosen…what can be better understood by playing the game?

There is a clear tendancy too focus on aspects that can be clearly and relatively accurately modelled in the statistical modelling process that is a set of wargame rules.

I agree. Part of that is the fact that a simulation is a procedural system and statistical modeling is a major factor.

Equally well other aspects cannot be acurately modelled as their are simply too many variables for a set of pen and paper rule mechanisms to cover.\This is where the "fudge" factor has to be applied to best of the designers knowledge and belief, or what has been described above as a set of prejudices.

I agree again. Every simulation medium, whether tabletop wargames or computers, have their intrisic limitations as well as unique benefits in simulating.

However, Another limitation wargame designers in our hobby face is lack of knowledge concerning simulation methodology. So, at the moment, for many folks it is difficult to tell which simulation limits are intrisic to tabletop wargames and which are simply the limited knowledge of the designers--such as which design factors have to be 'fudged' and which are a simple deficiency in design options because the designers don't know how simulations work and the current methodologies.

For instance, all simulation designers face that 'too many variables' for the system to handle--regardless of the system. And simulation designers everywhere often resort to that 'fudge' factor [at least in the way you are using the term…to fill in with a guess where there isn't enough data or because of system limitations. They call them a "WAG" or Wild-Assed Guess.

The difference is that the simulation designer has ways to test whether that WAG is a successful guess or not, whether it succeeds in simulating what needs to be simulated. With wargame designers, they just fudge and the gamer gets to guess how that helps or hinders that a 'better understanding' of history, assuming they even know where the designer fudged.

In the end, the wargame system either does what it was designed to, or it doesn't. Wargame rules are a finite set of procedures designed to do finite things through game play. How that play relates to history is also a finite question with a specific, finite answer. There are a variety of specific methods for answering that question--those eight tests being one set.

Sorry your bored. It is far more fun actually testing the simulations to see if they work than it is to talk about it.

Bill H.

Saying that a game can't bee a simulation is a testimony of one's ignorance and possibly cover up of lack courage in research.

The thousands of military professionals who constantly test, rehash, explore both past and future tactics through various types of games (boards, staff, computer, even miniatures) would testimony of the contrary.

What makes a game a better simulation for example: the use of proper tactics and "the feelings" of the period of the actual problems of your historical or real life counterparts.
it is a game as you can stop and start over again or say: nobody gets hurt etc.
What makes a good game out of a simulation: mechanics that get it moving, that make it fun and challenging.

Remember the wise words uttered in Oliver Stone's "Gardens of stone"
"Opinions are like s, everybody's entitled to one".

Jc